PATHWAYS
FOR DISASTER
RECOVERY AND
RESILIENCE
The Role of
Ongoing Social
Assistance and
the Poverty
Benefit Scheme
in Fiji After
Tropical Cyclone
Winston
This work is a product of the staff of The World Bank with external contributions and funded by the Australian Government through the
Department of Foreign Affairs and Trade (DFAT) and the Global Facility for Disaster Reduction and Recovery (GFDRR). The findings, analysis and
conclusions expressed in this document do not necessarily reflect the views of any individual partner organization of The World Bank, its Board
of Directors, or the governments they represent.
Although The World Bank, DFAT and GFDRR make reasonable efforts to ensure all the information presented in this document is correct, its
accuracy and integrity cannot be guaranteed. Use of any data or information from this document is at the user’s own risk and under no
circumstances shall the World Bank, GFDRR or any of its partners be liable for any loss, damage, liability or expense incurred or suffered which is
claimed to result from reliance on the data contained in this document. The boundaries, colors, denomination, and other information shown in
any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement
or acceptance of such boundaries.
Copyright Statement:
The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of
applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will
normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please
send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone
978-750-8400, fax 978-750-4470, http://www.copyright.com/.
All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H
Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail: pubrights@worldbank.org.
Photography: © World Bank/Vlad Sokhin
Please cite this work as follows: Pazmino, Nathalie, Kenia Parsons, and Alexandra Muñoz. (2022). The Role of Ongoing Social Assistance and the
Poverty Benefit Scheme in Fiji After Tropical Cyclone Winston. Washington, D.C.: World Bank.
Acknowledgements
This report was prepared by Nathalie Pazmino, Social Protection Specialist; Kenia Parsons, Social Protection
Specialist; Alexandra Muñoz, Social Protection and Jobs Consultant, with coordination from Yasser
El-Gammal and Philip O’Keefe, and contributions from Dung Doan, Jesse Doyle, Gracia Hadiwidjaja,
Sandor Karacsony, Yasuhiro Kawasoe, Aisha Mansur, Wiebke Stein and Vika Waradi. The team would
like to acknowledge the Social Protection team of the Department of Foreign Affairs and Trade (DFAT),
including Lisa Hannigan, Director, and Jacqui Powell, Assistant Director, who provided input and
guidance during the course of this report. We are grateful for the comments provided by internal and
external reviewers. Field work, data collection, and initial analysis were conducted by Tebbutt Research.
The report was edited by Angela Takats. Laframboise Design provided graphic design services.
The team would like to extend our sincere gratitude to the respondents who graciously and patiently
participated in this survey, sharing their stories of how the Poverty Benefit Scheme (PBS) has impacted their
lives and the lives of those they live with. Lastly, we would like to express our appreciation to DFAT
(Australian Government), USAID (U.S. Agency for International Development), and GFDRR (Global Facility
for Disaster Reduction and Recovery) for supporting this relevant research and the Ministry of Women,
Children and Poverty Alleviation (MWCPA) (Government of Fiji) for their contributions to this paper and
ongoing commitment towards social protection in Fiji.
1
Contents
Acronyms 3
Executive Summary 4
1. Introduction 8
2. Detailed Findings 11
Background: Social Assistance Programs and Methods 11
Fiji’s Social Assistance Beneficiaries: Experiences With and
Allocation of Benefits 14
Assessing Economic Recovery After TC Winston 16
Perceptions of Economic Resilience to Shocks 22
Housing and Reconstruction 25
The Role of Di erent Forms of Social Assistance and Support
for Immediate and Long-Term Recovery 27
Women’s Role in Recovery E orts 29
Preparedness in the Event of a Future Disaster 31
3. Key Findings 35
4. Recommendations 37
Annexes 39
Annex 1. Social Protection Schemes 39
Annex 2. Business Process Review (BPR) 40
Annex 3. Methodology of the Survey and
Focus Group Discussions 41
The Survey 41
Focus Group Discussions 42
Survey Sample Frame 43
Summary Statistics 43
Note on the Analysis 44
Study Limitations 44
Annex 4. Areas for Further Research 44
References 45
2
Acronyms
ALMP Active Labor Market Programs
BPR Business Process Review
CPA Care and Protection Allowance
DFAT Department of Foreign A airs and Trade
DAS Disability Allowance Scheme
DSW Department of Social Welfare
FGD Focus Group Discussions
FNPF Fiji National Provident Fund
FVP Food Voucher Program
HIES Household Income and Expenditure Survey
HH Household
MIS Management Information Systems
MWCPA Ministry of Women, Children and Poverty Alleviation
NEC National Employment Centre
PBS Poverty Benefit Scheme
PICs Pacific Island Countries
PMT Proxy Means Testing
SOP Standard Operating Procedure
SPS Social Pension Scheme
TC Tropical Cyclone
3
EXECUTIVE
SUMMARY
The study aims to better understand the The study presents the situation as of
role of ongoing cash transfers to support December 2019 (prior to COVID-19), which
recovery and resilience e orts in Fiji, prior continues to be valid considering the country
to COVID-19. It involved a review of Fiji’s is facing additional economic hardships.
flagship social assistance program, the Poverty Furthermore, the Fijian Government has
Benefit Scheme (PBS). The study included a adopted a new social assistance policy,
qualitative survey, focus group discussions adding relevance to these findings within a
(FGDs), and a business process review and context of social assistance reform and
assessment of the Department of Social investment in adaptive social protection.
Welfare’s (DSW) IT/database. It follows a 2016
impact evaluation of top-up transfers provided The objectives of the study included: (i)
after Tropical Cyclone (TC) Winston. examining the ability of Fijians living in the
most a ected areas to restore their livelihoods
Six years after TC Winston, Fiji has been and well-being in the aftermath of TC Winston;
facing the combined socioeconomic impacts (ii) assessing the role of Government social
of COVID- 19 and two Category 5 TCs in assistance (cash) benefits – the PBS – in
2020 (TC Harold and TC Yasa), as well as one contributing to recovery e orts; and (iii)
Category 3 TC in 2021 (TC Ana). Poor and examining the e orts undertaken at the
vulnerable Fijians are still struggling to restore household level to prepare for the possibility
livelihoods lost to TC Winston and they of future serious weather-related events.
generally remain disproportionately vulnerable
– both economically and socially – to the
e ects of other serious natural disasters.
4
February March April June … October December 2020-2021
2016 2016 2016 2016 2019 2019
TC Winston Social 75% of HHs Impact Business Survey and TC Harold Jan
assistance have evaluation process FGDs 20
top-up accessed of top-up … review TC Yasa Dec 20
payments transfer transfers TC Ana Jan 21
made payments conducted COVID-19
The study draws on the views of individuals economic context, with Fiji’s GDP growth
receiving PBS benefits (PBS beneficiaries), as rate at 0.4 percent in 2019, during the time
well as those who applied for PBS benefits of the survey.4 Furthermore, because of the
but were assessed as ineligible (non-PBS economic impact of the COVID-19 pandemic
bene ficiaries). There was an equal represen- and the recent TCs, the economy contracted
tation of male and female respondents in each by 15.7 percent in 2020, representing the
group from Ba, Rakiri, Tavua, and Korovou most severe contraction in Fiji’s history
districts.1 The site locations encompassed the (World Bank, 2022). The econo mic contraction
majority of TC Winston-affected households. and decline in employment opportunities
was mostly felt by vulnerable households,
Key Findings particularly those who rely on income from
Ongoing social assistance benefits are impor low-skilled labor. Employment opportunities
tant for household resilience and recovery were specifically constrained for Fijian women
from large shocks such as TC Winston and aged 15 to 64 years old, with female labor
other household shocks. The study found force participation at 45.5 percent compared
that the perception of the importance of social to 82.5 percent for that of males.5
assistance on long-term recovery seemed to
be influenced by household participation in
more than one program2. In addition, Help for
Homes cash grants3 were the preferred option This is a qualitative study that uses a quantitative model for the
1.
selection of the qualitative sampling frame in order to identify groups
of the non-PBS beneficiaries interviewed who with similar characteristics. Eligibility for PBS benefits is determined
based on proxy means test scores, which establish the eligibility
did not receive any social protection benefit. threshold. The proxy means test uses observable characteristics
All focus group participants were grateful for of the household or its members to estimate their incomes or
consumption, when other income data (salary slips, tax returns) are
ongoing cash benefits (the PBS, the Social unavailable or unreliable. As the name suggests, the obtained income
estimates are in turn used for means-testing, that is to determine
Pension Scheme (SPS), the Care and Protec- whether a household, or an individual, is eligible for support from a
tion Allowance (CPA), the Disability Allowance social protection program. Both groups were selected based on their
comparable proxy means test scores, were similarly affected by TC
Scheme (DAS), or food vouchers) they receive Winston, and have (i) similar land type and dwelling ownership status,
(ii) quality of dwelling/living conditions, (iii) possession of durable
from the Government of Fiji. goods, and (iv) household demographics. Therefore, the PBS and
non-PBS beneficiary groups remain comparable.
Despite having made significant progress in 2.
The findings are limited to the benefits and support received by those
rebuilding their personal economic security, surveyed and their ability to recall this information.
3.
Government-provided vouchers for the repair/reconstruction of
local infrastructure, and homes, those surveyed homes damaged in TC Winston.
reported continued economic hardship. This 4.
World Bank, World Development Indicators.
hardship is compounded by a difficult macro- 5.
HIES data (2019).
5
Both PBS beneficiaries and non-PBS benefi- opportunities, this finding indicates the impor-
ciaries (the near poor) described themselves tance of the PBS to deliver support to house-
as experiencing economic hardship (that is, holds in times of crisis.
unable to meet their basic food and non-food
Female beneficiaries (26 percent) were twice
needs), with 86 percent of PBS beneficiaries
as likely as male beneficiaries (13 percent) to
and 83 percent of non-PBS beneficiaries
have reported the PBS as their main source
saying they were among those experiencing
of household income.6 Some women reported
the highest levels of hardship compared to
using social assistance benefits to invest in
the rest of their village. A significant number
business ventures, providing a level of financial
of low-income Fijians described themselves
independence and income despite disparities
as worse-o financially than they were before
in gender roles and employment opportunities.
TC Winston occurred. Roughly half (47 percent)
of those surveyed in a ected communities Many poor and near poor Fijians remain
reported that they had less money (29 percent disproportionately vulnerable to the e ects
‘somewhat less’ and 18 percent ‘much less’) to of another serious disaster, with 84 percent
pay for things like food, housing, and electricity of respondents believing they personally
than they did before TC Winston. Employment will be seriously a ected by another natural
opportunities, as well as rates of employment disaster or storm in the next five years,
(and paid work) among those surveyed, including 37 percent who believe it is ‘very
declined significantly, and have yet to recover likely’. Neither PBS nor non-PBS beneficiaries
since TC Winston. Sources of income also indicated feeling prepared for future disaster
shifted dramatically. One-quarter of surveyed or shocks. Based on their past experiences,
respondents reported having worked o -farm only 21 percent of PBS beneficiaries perceived
before TC Winston; only 14 percent were their household to be ‘very prepared’ to
working at the time of the survey. Rates of recover within a reasonable time from a future
employment declined most among PBS ben- shock or disaster. Non-PBS male beneficiaries
eficiaries. Twenty-six percent of PBS benefi- (28 percent) were more than twice as likely as
ciaries worked pre-TC Winston; only 10 percent non-PBS female beneficiaries (12 percent) to
had paid work at the time of the survey. describe themselves as ‘very prepared’. Non-
Non-PBS beneficiaries experienced a notable, PBS female beneficiaries were also three times
albeit smaller, decline in paid work status as likely as their male counterparts to say they
pre- and post-TC Winston with 24 percent were ‘not prepared at all’. This highlights the
working pre-TC Winston and 19 percent importance of providing training and support
working at the time of the survey. to prepare and respond to future shocks.
Among PBS beneficiaries, the importance
of the PBS as households’ main source of
income quadrupled from 2016 to 2019;
9 percent of PBS beneficiaries indicated that
The household survey was conducted in December 2019 and covered
the PBS was their primary source of income
6.
a sample of current beneficiaries of the PBS (n=237, 119 male and 118
before TC Winston compared to 35 percent female) and non-PBS beneficiaries (n=209, 107 male and 102 female)
from a mix of rural and urban dwellers from Ba, Rakiri, Tavua, and
who indicated the PBS as their primary source Korovou districts. The sampling was done at an individual level and
although those surveyed provided information for their respective
of income at the time of the survey in 2019. households, it was not possible to obtain information for female- and
In an environment with lower employment male-headed households.
6
Ongoing cash assistance in the form of PBS
payments, in addition to Help for Homes
grants and access to free building supplies
(provided by government), enabled most
participants to rebuild village homes despite
their economic struggles. Furthermore, FGD
participants reported using social assistance
benefits to build back better, with stronger
materials that are more likely to withstand
future storms.
PBS benefits were almost entirely used to
pay for essential household expenses, with 4. Ensure groups that are particularly vulner-
food and household bills comprising the able to disasters (such as the PBS and
bulk of benefit expenditures. There were no near poor households, people living with
reported di erences in expenditure patterns disabilities, single mothers, older persons,
between male and female beneficiaries. and others) have timely access to informa-
tion, training and support to prepare and
Recommendations
respond to future shocks, and to develop
The report concludes with five key recommen- and implement disaster risk reduction
dations: plans in collaboration with other community
1. The government could consider building members and government agencies to
and improving existing systems to facilitate address specific hazards that the commu-
easy expansion of social assistance pro- nity will face.
grams (both vertically and horizontally) to 5. Identify opportunities to support women
provide predicted support to vulnerable and improve the inclusivity of disaster risk
non-PBS beneficiaries and top-ups to PBS reduction programming to involve women
beneficiaries. as agents of change.
2. Further analysis is needed to consider how The report that follows provides a more
social protection can be tailored (in normal detailed analysis and breakdown of the
times and in response to covariate shocks) results, including di erences in experiences
to address specific risks, vulnerabilities, and between Fijian men and women, PBS benefi-
inequalities for women and marginalised ciaries and non-PBS beneficiaries, in addition
groups, in order to improve their resilience to experiences with social assistance programs
to disasters. and top-up disaster payments. More detailed
3. Assisting vulnerable households, particu- recommendations to increase the e ective-
larly female beneficiaries, to diversify their ness of ongoing social protection to support
income through access to training and fi- recovery and increase resilience are also pre-
nance could be achieved by establishing sented.
linkages and referral systems to connect
beneficiaries of Fijian social protection
programs with other service providers in
their regions such as employment programs
and financial inclusion services.
7
1. INTRODUCTION
Fiji is one of the largest and most developed speeds reaching 233 km/hour and wind gusts
of the Pacific Island Countries (PICs). It is peaking at around 306 km/hour, making it
an archipelago of 332 islands (of which one of the most powerful cyclones ever
approximately 110 are inhabited), and spreads recorded in the Southern Hemisphere
over a land area of approximately 18,300 km2. (Government of Fiji, 2016). The Eastern
The country has a population of approximately Division of Fiji was the first to be hit by TC
890,000, with the majority of people residing Winston, which severely damaged the Islands
primarily on the two largest islands of Viti Levu of Taveuni, Koro and Ovalau. It also damaged
and Vanua Levu. other islands within its path before reaching
its peak strength and making landfall on Viti
External shocks, such as natural disasters
Levu. In addition to the extreme wind speeds,
due to climate change, have been one of the
storm surges led to flooding on many islands,
main challenges in the country. According to
and in some cases, inundated areas almost
the 2020 World Risk Report, Fiji is ranked as
200 meters inland.
the country facing the 15th highest level of
disaster risk in the world. It is also ranked as TC Winston caused widespread damage and
highly susceptible to climate change impacts. destruction and reportedly impacted over
Between 1972 and 2009, the country 540,000 people, or 60 percent of the total
reported 124 natural disasters; tropical cyclones population of Fiji. It caused 44 fatalities and
accounted for 50 percent of these disasters, around 40,000 people required immediate
followed by floods (33 percent) and earthquakes assistance. Over 30,000 houses, 500 schools,
(8 percent) (Burton, 2015). In 2016, TC Winston and 88 health clinics and medical facilities
caused an estimated F$2 billion of damage, or were damaged or destroyed. Approximately
25 percent of GDP (Mansur et al., 2017). 80 percent of the nation’s population lost
This has been exacerbated by the country’s power, including the entire island of Vanua
vulnerability to climate change, which is causing Levu. In terms of livelihoods a ected by
an additional estimated loss of 2.5 percent of income loss, 57 percent were in the agricul-
GDP every year. tural sector, 17 percent in commerce, 10 percent
in manufacturing, 8 percent in tourism, and
TC Winston, an extremely destructive Category
8 percent in transportation.
5 cyclone, struck Fiji on 20th February 2016.
It was the first Category 5 cyclone to directly
impact Fiji, with maximum average wind
8
their regular social assistance payment for
three months. Under the SPS, 17,782 pensioners
over the age of 68 received an additional
F$300. Under the CPA, 3,313 families received
a total of F$300 each (Mansur et al., 2017).
The government also rolled out a new
housing initiative called Help for Homes. The
F$70 million (US$32.6 million) program began
on April 9, 2016, to assist families to rebuild
homes that were damaged or destroyed. A
total of 30,369 homes were reported
damaged or destroyed by TC Winston, and
under this program the government provided
a ected households with vouchers (electronic
cards) for housing rehabilitation/reconstruc-
tion. The program targeted households
with an annual income of under F$50,000
(US$24,000), who experienced housing damage.
The three categories of benefits provided
were: F$1,500 (US$717) for houses with partial
roofing damage; F$3,000 (US$1,434) for
houses with a serious roofing damage; and
F$7,000 (US$3,345) for almost/completely
demolished households. Those living in infor-
mal settlements in a ected areas were also
The Government of Fiji decided to use its eligible to receive F$1,500 (US$750) (Mansur
existing social protection system to provide et al., 2017). In addition, applicants who had
additional assistance to the most vulnerable, already repaired their homes could request
as a key component of its disaster response, refunds but needed to provide evidence.
which provides regular cash and food voucher
This study contributes to lessons learned
payments to vulnerable families. Fiji’s Depart-
from previous studies in order to strengthen
ment of Social Welfare manages several social
the design of shock-responsive social
protection programs including the Poverty
protection in the Pacific region. This includes
Benefit Scheme (PBS), the Social Pension
improved information and knowledge on
Scheme (SPS), the Care and Protection
poverty and vulnerability in order to better
Allowance (CPA), and the Disability Allowance
protect incomes and livelihoods of the most
Scheme (DAS).7 Four weeks into the disaster,
vulnerable and to strengthen Pacific Island
the Government disbursed F$19.9 million
countries’ preparedness and resilience to natural
(US$9.4 million) using three of the existing
disasters and climate change in the region.
social assistance schemes to e ciently
disburse cash to vulnerable groups and inject
much needed cash into the economy. Under
the PBS, 22,802 households were paid a lump The Disability Allowance Scheme was introduced after TC Winston
7.
and therefore beneficiaries under this scheme did not receive the
sum of F$600, or the equivalent of a tripling of top-up payments.
9
In 2016, an impact evaluation of the top-up
transfers and Help for Homes grants follow- • Households acted rationally in their
ing TC Winston was conducted by the World spending patterns and the top-up pay-
Bank in conjunction with the Fiji Bureau of ments were used on essential items.
Statistics and the Australian Department of
• Knowledge of the top-up transfers was
Foreign A airs and Trade (DFAT). This shock-
very low amongst the recipients, with
responsive social protection intervention, in
the majority of recipients unaware that
the form of top-up transfers to beneficiaries,
they would be receiving the payment.
was found to be an e ective immediate
Source: Mansur et al., 2017.
response following TC Winston (Mansur et al.,
2017). In 2019, however, TC Winston recovery
e orts were ongoing in some of the hardest This study seeks to understand the role of
hit communities. Furthermore, a challenging ongoing social transfers during and
macroeconomic context has served as an following TC Winston in sustaining recovery
obstacle to recovery. Given the continued need e orts and growing household resiliency
for learning from implementation, particularly to future shocks, particularly within a
in light of the renewed social assistance complicated macroeconomic context. While
policy as well as new investments in adaptive the 2016 evaluation investigated the impacts
social protection, this study presents findings of cash payments in the immediate aftermath
at a critical moment of analysis and action. of the disaster, this follow up study focuses on
particularly assessing:
• The role of Government-provided ongoing
Impact Evaluation of Top-up Payments
social assistance (cash transfer) benefits in
2016: Key Findings
contributing to recovery e orts.
• The disaster-responsive social protection
• The ability of Fijians living in the most a ected
intervention, in the form of top-up
areas to restore their livelihoods and well-
transfers to beneficiaries, was found to
being in the aftermath of TC Winston.
be an e ective response following TC
Winston. • E orts undertaken at the household level to
prepare for the possibility of future serious
• In-kind humanitarian assistance provided
weather-related events.
e ective immediate support to all a ect-
ed households, and played a crucial The study defined household resilience as
role in sustaining families until markets to whether households felt they were more
were restored. prepared to recover from a future disaster.
• Both poor and near poor households
were similarly a ected by TC Winston,
and the households receiving the top-up
transfers were quicker to recover from
the disaster shocks.
10
2. DETAILED FINDINGS
Background: Social Assistance amount paid to caregivers ranges from
F$29 to F$127 per household per month.
Programs and Methods
DAS: provides a monthly benefit of F$90 to
The main building blocks of Fiji’s social
people with severe disabilities.
protection system are: (i) categorical- and
poverty-targeted social assistance benefits FVP: implemented to encourage pregnant
implemented by the MWCPA; (ii) social insur- mothers in rural settings to attend clinics in
ance (through a mandatory defined-contribu- the early stages of pregnancy, to avoid
tion pension scheme in the formal sector) complications during delivery and to mini-
implemented by the Fiji National Provident mize cases of child malnutrition. It gives
Fund (FNPF); and (iii) active labor market food vouchers worth F$50 a month for 10
programs (ALMPs) implemented by the National months depending on the number of
Employment Centre (NEC). months they visit health facilities.
Under the MWCPA, Fiji has one of the most In addition, the MWCPA is the agency
comprehensive social assistance system responsible for the implementation of the
among all PICs. The current core elements of National Gender Policy for Fiji, which aims to
the system are the PBS, SPS, CPA, DAS, and promote gender equity, equality, social
the Food Voucher Program (FVP) for Rural justice, and sustainable development. As a
Pregnant Mothers. response to high rates of Gender Based
Violence (GBV) in Fiji, the MWCPA has put in
PBS: targeted to poor families based on proxy
place the National Service Delivery Protocol
means testing (PMT) and covers approxi-
as a common set of principles and guidelines
mately 26,000 households (potentially
for service providers to abide by. To improve
112,000 indirect beneficiaries). Families
the impact of social assistance in Fiji, the
receive F$35 per adult and F$17 per child
MWCPA, Department of Social Welfare (DSW),
(below 17 years of age), with a maximum of
has been undertaking an ambitious set of social
F$127 per household per month.
assistance reforms over the past 10 years with
SPS: a social pension scheme for those the support of the World Bank and DFAT.
above 65 years of age not receiving gov-
ernment pensions, ex-serviceman funds or The Survey Respondents
the FNPF, or any other superannuation, A qualitative household survey was conduct-
with a benefit of F$100. PBS and SPS bene- ed in December 2019 and covered a sample
ficiaries also receive food vouchers worth of beneficiaries of the PBS (n=237, 119 male,
F$50 a month. and 118 female) and non-PBS beneficia-
CPA: a child benefit given to underprivi- ries (n=209, 107 male, and 102 female) in
leged children for care and protection. The 2019, i.e. at the time of the survey, and
from a mix of rural and urban dwellers from
11
Ba, Rakiri, Tavua, and Korovou districts. While this is a qualitative report, the team
For this study, PBS-beneficiaries are successful was careful to build the case for selection of
applicants of the PBS benefit, who self-registered individuals based on a quantitative model
as household heads and are in charge of (that is, PMT) and using a suitable sampling
receiving the benefit on behalf of the household. frame for this design (that is, the DSW PBS
While sampling was done on an individual registry database). The survey sample was
level and data was disaggregated based on delineated along the lines of beneficiary status
respondents’ characteristics, those surveyed in order to determine whether there were
provided information for their respective di erences in recovery and resilience over
households and the role of social assistance time between households that received
was analyzed at a household level. In other ongoing PBS benefits from the Government
words, the survey interviewed individuals (PBS of Fiji and those that did not. This was done in
and non-PBS beneficiaries) who reported order to understand the long-term role of
on their own perceptions of PBS benefits ongoing cash transfers beyond the immediate
for themselves and for their households. aftermath of TC Winston.
(See Annex 3 for detail sampling design). In a number of cases surveyed individuals
The non-PBS beneficiaries are individuals resided in households that received benefits
who applied for the PBS but were not from another government social assistance
assessed as eligible for PBS benefits because program. About 35 percent of PBS beneficiaries
they were slightly over the cut-o point of and 41 percent of non-PBS beneficiaries
the proxy means test. The concept of proxy personally received or said a household
means testing is usually understood as using member received the SPS, CPA, or DAS. Top
observable characteristics of the household up payments were also given to the beneficiaries
or its members to estimate their incomes, of these schemes.8 There were no statistically
based on consumption, when other, when significant di erences in household participa-
other income data (salary slips, tax returns) tion in each of the other social assistance
are unavailable or unreliable. As the name programs between the two groups, suggesting
suggests, the obtained income estimates are that a significant share of the non-PBS bene-
in turn used for means-testing, that is to ficiary group was also comprised of households
determine whether a household, or an with vulnerable and disadvantaged individuals.
individual, is eligible for support from a social Table 1A shows sex-disaggregated data
protection program. If their estimated income of beneficiary households in other social
falls below a certain threshold, they are assistance programs. Within the non-PBS
deemed eligible for support; if the income is beneficiary group, female respondents were
higher than the threshold, they are deemed significantly more likely to live in a household
ineligible. Eligibility is reviewed every three enrolled in the SPS program than their male
years. Non-PBS beneficiaries are defined counterparts (Table 1B).
by DSW as individuals living near or just
above this poverty threshold, having therefore
comparable characteristics with the PBS
beneficiary group.
8.
Except for the beneficiaries under the Disability Allowance Scheme, which had not been introduced at that time.
12
Table 1. Participation of household members in other There are a number of limitations that need
social protection programs
to be taken into account when interpreting
1A. PBS beneficiaries
the data and drawing conclusions. It is
Total Male Female important to note that the survey is based on
(N=119) (N=118) perceptions of PBS and non-PBS beneficiaries.
Only PBS 154 Likewise, the study relies on respondents to
(65%) 71 83
recall the period before, in the immediate
PBS and 51
SPS (22%) 29 22
aftermath, and following the crisis, which is
also subjective, especially since the survey took
PBS and
CPA 17 (7%) 9 8 place in December 2019 and the cyclone hit
PBS and
the island in February 2016. For instance, the
DAS 17 (7%) 12 5 study found an issue in the di erences in the
1B. Non-PBS beneficiaries recollection of respondents in terms of their
Total Male Female participation in the social assistance schemes
(N=107) (N=102) and information provided by the DSW. Indi-
No 124 viduals who participated in this study were
benefits9 (59%) 69 55
confirmed by DSW systems as having en-
SPS 45
(22%) 16 29
rolled in the PBS pre-TC Winston and having
CPA 23 received benefits continuously since that
(11%) 12 11 time. Interviews were limited to Eastern and
DAS 21 Northern parts of Viti Levu island. Findings
(10%) 13 8
should be interpreted in consideration of
these limitations.
9.
This subgroup did not receive any social protection benefit (that is, no PBS, SPS, CPA, or DAS), as opposed to some of the other social protection
benefits listed in this table, where overlap of benefits is possible
13
Fiji’s Social Assistance What Are Benefits Used For?
Beneficiaries: Experiences Survey respondents were asked the question
With and Allocation of Benefits ‘Thinking about your regular, monthly cash
benefits you receive from Government, do
Intra-household Decision Making you share your benefits with anyone or any
In terms of gender and decision making of organisation outside of your household?’ The
recipients surveyed, 72 percent of the PBS survey found that PBS benefits are largely
beneficiaries (57 percent male and 86 percent kept within the household and allocated to pay
female) indicated that they are personally and for household-related expenses (93 percent).
primarily responsible for deciding how their Only 7 percent of PBS beneficiaries shared
household will spend cash benefits they their benefits with an individual or organisation
receive from the Government of Fiji each outside of the home. This is similar to the results
month10 (Figure 1). This was echoed in the focus of the previous impact evaluation which found
group discussions11 with participants explaining that sharing of both PBS benefits and top-up
that each beneficiary usually decides on how transfers with other households was very low.
to use their own resources. Another 16 percent This suggests that, in times of disaster, assistance
indicated that they decide jointly with a spouse should be closely targeted to the most a ected
or partner how to spend social assistance households (Mansur et al., 2017).
money, and a further 8 percent of beneficiaries While the qualitative data found some
indicated that they make decisions jointly with indications of pressure to contribute to village/
various other members of their household. church obligations, most PBS beneficiaries
Male PBS beneficiaries indicated that they are surveyed and participants of the FGD reported
significantly more likely than female PBS no major pressure to share benefits. As reflected
beneficiaries to share decision making with in the survey data, PBS beneficiaries who
their spouse or partner. It is important to keep participated in focus groups did not feel pressured
in mind that only 35 percent of women who to share benefits outside the home. A Ba
participated in the survey reported being married woman explained, “I decide on my own who to
compared to 88 percent of men. share my monthly benefit with.” This was
echoed in the majority of focus group discus-
Figure 1. Decision making on PBS expenditure
for male and female beneficiaries sions. A handful of men, however, commented
that benefits are sometimes used to contribute
100% to village obligations or shared with the church.
90% One Rakiraki man raised that there were
80%
70% reports of local divisions between community
60% members who do and do not receive benefits,
50%
40%
saying, “If there is a village obligation, and we
30% do not give our share, people will point fingers
20%
at us and say, ‘If we can give, why can’t you?
10%
0% You have the government assisting you with
Own decision Couple decision
Male Female
10.
According to the PBS Operating Procedures (item 2.4), the PBS is paid to the head of the household, being male or female. However, a World Bank
analysis of the HIES 2019 demonstrates that the PBS benefit has equal gender coverage (49.4 percent of females and 50.6 percent of males, as
direct and indirect beneficiaries).
11.
See Annex 3 for detailed information on focus group discussions.
14
money’... Also, those that have nothing have Non-PBS beneficiaries spent a greater pro-
grown accustomed to borrowing, knowing portion of their monthly expenditure on home
that we have [our benefits].” repairs, while PBS beneficiaries prioritised
PBS beneficiaries acted rationally in their health-related expenses. PBS beneficiaries
spending patterns; PBS money was used were just as likely as non-PBS beneficiaries to
almost entirely to pay for essential household have received assistance through the Help for
needs, with food expenses comprising Homes program after TC Winston (70 percent
the bulk of benefit expenditures. More than among beneficiaries and 72 percent among
90 percent of PBS beneficiaries reported to non-PBS beneficiaries). Therefore, di erences
primarily using monthly benefits to pay for in the proportion of household expenditure
food. Benefits remaining after paying for food dedicated to home repairs may indicate that
were reported to be largely used to pay the combined benefits of PBS and Help for
for bills like phone, gas, and electricity Homes assistance allowed PBS beneficiaries to
(70 percent), and health-related expenses (54 cover home repair expenses more e ciently
percent). There were no di erences in the than non-PBS beneficiaries.
household expenditure priorities reported by
male and female respondents (Figure 3).
Figure 2. Top household expenditures12
Food
Food
Bills
Bills
Health
Health
Education
Education
Home repairs
Home repairs
Social obligations
Non-PBS beneficiaries
Social obligations
0 20 40 60 80 100 120 PBS beneficiaries
Non-PBS beneficiaries
0 20 40 60 80 100 120 PBS beneficiaries
Figure 3. Top household expenditures by male and female beneficiaries
100%
100%
90%
90%
80%
80%
70%
70%
60%
60%
50%
50%
40%
40%
30%
30%
20%
20% Men
10% Men
10%
0% Women
Women
0%
Food
Food Bills
Bills Health
Health Education Home
Education Home repairs Social
repairs Social
obligations
obligations
12.
All results displayed between PBS and non-PBS beneficiaries are within the margin of error. In addition, significance testing was run within questions
to determine significant di erences between subgroups in question responses (male and female). Significant di erences in the results indicated
above are di erences > the margin of error for the samples (+/-6 percent at the 95 percent confidence interval).
15
Assessing Economic Recovery For non-PBS beneficiaries, household
After TC Winston participation in other programs increased
respondents’ perception of being able to
Perceptions of Economic manage hardship most of the time. Although
Hardship13 perceptions of hardship were similar among
PBS beneficiaries and non-PBS beneficiaries,
Most Fijians living below or near the poverty individuals who did not benefit from any
level described their household as ‘experiencing social welfare scheme (PBS, SPS, CPA, DAS)
hardship’, including approximately two in were most likely (44 percent) to perceive their
five who described their household as among household as experiencing the highest level
those ‘experiencing the highest level of of hardship in their village. In comparison, this
hardship’ in their village. Perceptions of personal number dropped by 13 percentage points for
hardship were similar among PBS beneficia- those who received benefits from the CPA,
ries and non-PBS beneficiaries. Among the SPS, or DAS. This indicates the value of having
surveyed households, 86 percent of PBS and access to a broad mix of safety net programs
83 percent non-PBS beneficiaries reported that to alleviate economic hardship, and may also
their household was among those currently indicate the presence of exclusion errors.
experiencing ‘the highest level of economic
hardship in their village’ or ‘some level of
hardship’. Only 1 percent of PBS beneficiaries
and 4 percent of non-PBS beneficiaries
considered themselves among the ‘better o
households in their village’ (Figure 4).
Figure 4. Perceived economic status
within the community
100% 1% 4%
13%
12%
80%
60%
86% 83%
40%
20%
0%
PBS Non-PBS
Among the most disadvantaged Managing
13.
In the Pacific, people living in economic hardship are those unable to meet their basic food and non-food needs.
16
Table 2. Participation in social protection programs and perceived economic status within the community
Net: Highest and
some level of
hardship Manages
PBS beneficiaries 86% (37% highest) 13% 1%
Non-PBS beneficiaries 83% (39%) 12% 3%
PBS only 88% (35%) 10% 1%
PBS + CPA, SPS, or DAS 81% (40%) 17% 0%
CPA, SPS, or DAS only 79% (31%) 15% 4%
No benefits 86% (44%) 10% 2%
Reported Rates of Employment Winston compared to 22 percent of men
working at the time of the survey; 15 percent
Among respondents, rates of employment
of women worked outside of home pre-TC
have declined significantly since TC Winston.
Winston compared to just 5 percent at the
Survey findings indicated that while 25 percent
time of the survey. Comparisons of PBS and
of total survey respondents reported having
non-PBS groups revealed that rates of
worked before TC Winston occurred, only
employment among non-PBS males stayed
14 percent reported working at the time of the
relatively similar pre- and post-TC Winston,
survey in 2019. Rates of employment declined
while declining significantly among males
most among PBS beneficiaries in a ected
who received PBS (Figure 5). Similarly, rates
areas, though non-PBS beneficiaries also saw
of employment for women, which were
devastating impacts to their ability to work as
already lower for female respondents, fell by
a result of TC Winston. While 26 percent of
66 percent for PBS beneficiaries. Inequalities
PBS beneficiaries worked pre-TC Winston, only
in the labor market in Fiji are particularly
10 percent reported working at the time of the
challenging for women aged 15 to 64 years
survey. Non-PBS beneficiaries experienced a
old, with female labor force participation
notable, albeit smaller, decline in work status
considerably lower than that of males, at
pre- and post-TC Winston, with 24 percent
45.5 percent compared to 82.5 percent.14
working pre-TC Winston and 19 percent
A lack of access to quality childcare support
working at the time of the survey.
poses a key constraint on women’s participation
In general, women were significantly less and re-entry into the workforce (IFC, 2019).
likely to work outside home than their male Given that women are overrepresented in
counterparts. While rates of employment some of the employment sectors most
declined significantly among both men and impacted by COVID-19 (Dornan et al., 2020),
women post-TC Winston, only 5 percent of this disparity is likely to be exacerbated. This
women in a ected areas reported working highlights the importance of the PBS as a
outside the home at the time of the survey. social safety net in order to avoid the further
Thirty-five percent of men worked pre-TC decline into poverty of vulnerable groups,
such as women.
14.
HIES data (2019).
17
Figure 5. Level of employment pre- and post-TC Winston
PBS beneficiaries Non-PBS beneficiaries
40%
35%
30%
25%
20%
15%
10%
5%
0%
MALE FEMALE MALE FEMALE
Working pre-TC Winston Working pre-TC Winston
Working post-TC Winston Working post-TC Winston
The study indicated that sources of income of PBS beneficiaries reported the PBS as
have shifted in the years since TC Winston, their primary source of income before TC
with significantly fewer households bene- Winston compared to 35 percent who said
fitting from paid work opportunities and since TC Winston they consider the PBS
more families benefiting from some form as their primary source of income in 2019
of social welfare benefit. This is likely due in (Figure 6). Among the non-PBS beneficiaries,
part to the overall macroeconomic situation, the sources of income from paid work also
in which low GDP growth and constraints to declined, though not to the same extent.
employment have made it more di cult Given the negative impact of COVID-19
for vulnerable households. This can be said on economic outcomes throughout the
for both the PBS beneficiary and non-PBS region, income from employment is likely to
beneficiary samples. Among the PBS benefi- fall further across both groups, highlighting
ciaries, the importance of PBS benefits as the potential need for both vertical and
the households’ main source of income horizontal expansion of benefits.
quadrupled from 2016 to 2019. Just 9 percent
Figure 6. Main sources of income pre- and post-TC Winston
PBS beneficiaries Non-PBS beneficiaries
PBS Other
Remittances SPS
Wages and salary - Wages and salary -
ongoing work ongoing work
Casual work Casual work
Informal business
Informal business
Farming
Farming
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45%
Post-TC Winston Pre-TC Winston Post-TC Winston Pre-TC Winston
18
TC Winston damaged agricultural livelihoods in Most survey respondents reported that
the survey areas and, as of December 2019, the they were still struggling to rebuild their
sector has not fully recovered. Respondents livelihoods since TC Winston and were
from both groups reported that household worse o financially at the time of the survey
income tied to farming declined in the almost than they were before the cyclone (Table
four years since TC Winston occurred. PBS 4). Male PBS beneficiaries (48 percent) and
beneficiaries indicated that they were currently female PBS beneficiaries (47 percent) were
9 percentage points (51 percent to 42 equally as likely to report having less money
percent) less likely to consider farming as one to pay for essential household items like
of their top sources of income than they were food, electricity, or housing at the time of
before TC Winston occurred. the survey compared to before TC Winston.
Among the non-PBS sample, 25 percent of
In keeping with disparities in rates of
male respondents and 24 percent of female
employment between men and women,
respondents reported having more money
sources of income di ered considerably
at the time of the survey to pay for essential
between men and women in a ected areas
household items than before TC Winston,
in 2019 (Table 3). Considering the combined while 51 percent of male respondents and
samples of PBS beneficiaries and non-PBS 44 percent of female respondents reported
beneficiaries, women (26 percent) were twice that the household had less money than
as likely as men (13 percent) to use the before the cyclone.
PBS benefit as their main source of income,
Table 4. Perceptions of income variability
whereas men were much more likely to derive
income from farming (46 percent) than Net more Net same Net less
women (23 percent). Moreover, women (10 money money
percent) were three times as likely as men (3 PBS men 19% 31% 48%
percent) to garner household income from PBS women 17% 34% 47%
Non-PBS men 25% 23% 51%
informal business activities like handicrafts
Non-PBS women 24% 30% 44%
and selling food. Comparing PBS and
non-PBS groups, women were more likely Levels of periodic di culty in a ording
than men to have social assistance benefits essential household items were high before
(mostly PBS and SPS) as their main source of TC Winston occurred and this was still the
household income. case in 2019. Most respondents reported
Table 3. Main sources of household income by sex having experienced periodic di culty in
Main source of income Male Female
a ording food, healthcare, household bills,
home maintenance and repairs, and school-
No source of income 3% 5%
ing both in the months before TC Winston
Farming 46% 23%
and at the time of the survey (Figures
Informal business 3% 10%
7A–7E). A Rakiraki man who participated in
Casual work 11% 8%
focus groups explained, “There are things
Wages and salary – ongoing work 7% 10%
that we struggle to pay for still – medical
Remittances 0 1%
expenses, furniture, transportation, and
Poverty Benefit Scheme 13% 26%
stationery for my children’s [schooling].”
Social Pension Scheme 4% 7%
Care and Protection Allowance 1% 0
Disability Allowance Scheme 4% 2%
Note: Color blocked cells indicate significant di erence between male
and female.
19
pre- and post-TC Winston
7A. Food 7B. Healthcare
100% 100%
80% 95%
60% 90%
40% 85%
20% 80%
0% 75%
PBS Non-PBS PBS Non-PBS
Pre-TC Winston % Yes Post-TC Winston % Yes Pre-TC Winston % Yes Post-TC Winston % Yes
7C. Bills 7D. Schooling
100% 100%
80% 80%
60% 60%
40% 40%
20% 20%
0% 0%
PBS Non-PBS PBS Non-PBS
Pre-TC Winston % Yes Post-TC Winston % Yes Pre-TC Winston % Yes Post-TC Winston % Yes
7E. Repairs PBS beneficiaries were 7 percentage points
less likely to report having had di culty
100% a ording food at the time of the survey
compared to the same period leading up to
80% TC Winston. As shown in Figure 8, this change
was largely driven by female beneficiaries.
60%
Non-PBS beneficiaries were 8 percentage
points more likely than PBS beneficiaries to
40%
report having di culty getting and paying
20% for food in recent months. On average, PBS
beneficiaries were less likely to report having
0% di culty in accessing and purchasing food,
PBS Non-PBS as opposed to non-PBS beneficiaries. Both
groups were still struggling to pay for home
Pre-TC Winston % Yes Post-TC Winston % Yes
repairs at the time of the survey.
20
Both PBS and non-PBS beneficiaries faced their female counterparts to report having
more difficulties getting and paying for experienced difficulties affording schooling
healthcare at the time of the survey, in for their children. In terms of getting and
comparison to the period before TC Winston. paying for food, women from households
Considering that the cash transfers received receiving PBS benefits were better off at the
by PBS beneficiaries had been reducing their time of the survey by 10 percentage points,
difficulties in paying for other bills and services, while women from non-PBS households
the increased difficulties accessing and were 3 percentage points worse off.
paying for healthcare suggests the existence Non-PBS beneficiary women were more
of supply side constraints that have not been likely to have difficulty paying household
resolved.15 bills post-TC Winston by 9 percentage
points. Both PBS and non-PBS households
The study found that perceptions of
reported having difficulties paying for
hardship pre- and post-TC Winston differed
healthcare and home maintenance post-TC
between male and female respondents
Winston, and this did not differ between
from PBS and non-PBS households (Figure 8).
males and females.
Within samples, men were more likely than
15.
Although Fiji has 19 subdivisional hospitals, rural areas are covered by nursing stations that are typically staffed by one registered nurse and cater
for a catchment population of up to 5000 people (Asante et al., 2017).
21
Figure 8. Change in hardship pre- and post-TC Winston
10
8
6
4
2
0
-2
-4
-6 Male PBS
Female PBS
-8 Male non-PBS
-10 Female non-PBS
Getting and Getting and paying Paying bills Paying for Paying for home
paying for food for healthcare schooling maintenance or
repairs
Perceptions of Economic and often require larger e orts to rebuild
infrastructure, such as natural disasters. While
Resilience to Shocks
households may be able to rely on informal
Fijian households have been exposed to social safety nets to overcome idiosyncratic
multiple hazards and shocks that have shocks, recovery from covariate shocks often
impacted their recovery e orts. In December underlines the importance of formal social
2019, 40 percent of PBS beneficiaries and assistance benefits. The most frequently repor-
35 percent of non-PBS beneficiaries reported ted shocks included flooding (21 percent), a
having experienced another catastrophic event death, illness, or dissolution of marriage (10
or shock, post-TC Winston. These shocks percent), or unemployment (6 percent). There
included both idiosyncratic shocks – a parti- were no di erences in the reported exposure to
cular experience of one household, such as catastrophic events or shocks post-TC Winston
death, injury, or unemployment – and cova- between male and female respondents.
riate shocks that impact the whole community
22
Recounting the Trauma of TC Winston
“Winston took away much of our source of income. We were pushed into
poverty due to the tropical cyclone and floods.” – Ba woman
“We had to start from scratch. The village looked like a war zone after the
cyclone.” – Korovou woman
“Winston destroyed my vegetables and root crops plantation. I had to start
from scratch, and it took me three years to recover these things back.” –
Korovou woman
“Life has not fully recovered like the way it was before Winston.” – Rakiraki woman
Focus group participants struggled to retain their composure while recount-
ing their experiences during and immediately after TC Winston. Discussions
brought back memories of trauma, including the deaths of villagers. Male and
female participants were visibly upset as they recounted the mental, physical,
and emotional toll wrought by the storm. Some described the dissolution of
marital relationships in the aftermath of TC Winston due to stress. After TC
Winston, the government provided psychological counseling to villagers, but
services were short-lived, and focus group discussions illustrated the need for
additional counseling support.
While economic recovery and resiliency levels differed considerably between
villages where focus groups took place, life for most was just starting to return
to normal after more than three years spent struggling to rebuild and recoup
livelihoods lost.
“It took us three years to recover our lives back.” – Ba man
“I can say that I have slowly recovered but it took me years, not months.” – Ba
woman
“The entire community saw their homes and livelihoods destroyed or severely
compromised.” – Ba man “We had to look for other sources of income for
survival.” – Rakiraki man
While some livelihoods in 2019 were returning to pre-TC Winston levels, oth-
ers (outside of Ba) described themselves as worse off financially than they
were pre-TC Winston. This may suggest that a review by DSW of the current
PBS benefit amount could be useful.
“There is no change at all. We are still struggling every day since Cyclone
Winston.” – Tavua man
“I would say that my economic situation is worse than it was before Winston.
The cost of living is very high...Some of us cannot afford the cost of food
items.” – Tavua man
23
“We still struggle to meet our everyday living, and the assistance from government
has assisted us in taking care of some things.” – Tavua man
Residents of Navala Village, in the highlands of Ba, appeared more resilient
economically compared to other villages where focus groups were held. It has
taken years for villagers to recover their livelihoods, but economic outlooks have
improved in their village. Male participants were able to increase farm productiv-
ity through regrowth of Kava, and supplement individual incomes from tourism
profits.
“Our economic situation is now stable. It is better than it was straight after
Winston struck...When Winston happened, we had to start again from the bottom
and slowly work our way back up.” – Ba man
“There are still bits and pieces missing in the puzzle of putting our lives back in
order but most of it has [worked out].” – Ba man
“Our village is known for its thatched houses and tourists visit the village to catch
a glimpse of the real Fijian way of living. We charge each tourist F$25.16 Funds are
kept in a village trust fund, and if anyone needs medical care or there are
problems with our roads...we step in [as a community] to assist.” – Ba woman.
In addition to tourism dollars, Ba, Rakiraki, and Tavua participants garnered in-
come by leasing land rights and through paid work inside and outside their vil-
lages in factories, on sugar cane farms, or through paid domestic work. Korovou
residents, by comparison, seemed to have fewer sources of income to fall back
on – other than benefit schemes – since TC Winston. Some residents diversified
sources of income by using cash benefits to establish small businesses such as
canteens, subsistence farms, tailors, and money lending businesses, or by rein-
vesting funds to improve farming practices.
“[I have used my benefits to buy] manure and weed chemicals for my crops, to
market my produce and generate more income.” – Korovou woman
16.
US$12 (exchange rate FJD/USD 0.48 as of April 7, 2021).
24
Housing and Reconstruction PBS and non-PBS beneficiaries reported very
similar experiences with rebuilding e orts.
According to the Government of Fiji, TC
About 30 percent of survey respondents
Winston obliterated more than 9,000 houses,
whose homes were damaged have been unable
substantially damaged almost 17,000 houses,
to finish repairing all of the damages incurred
and rendered minor damage to another
by TC Winston (31 percent PBS beneficiaries,
29.000 (NationalDisaster Management O ce,
30 percent non-PBS beneficiaries) and a small
2016). In this study, about 35 percent of PBS
number (3 percent PBS beneficiaries, 4 percent
beneficiaries and 32 percent of non-PBS
non-PBS beneficiaries) had to move to a
beneficiaries reported their homes su ered
di erent home entirely because their previous
substantial damage (that is, more than half
home was beyond repair.17 One-fifth of male
their home needed to be repaired or rebuilt).
PBS and non-PBS respondents reported that
Another 19 percent of PBS beneficiaries and
their homes were stronger at the time of the
18 percent of non-PBS beneficiaries reported
survey than before TC Winston. This percentage
that their homes were completely destroyed
was lower for female respondents, with PBS
by TC Winston. Upwards of two in five
female respondents more likely to report
houses experienced minor (18 percent for
stronger homes (22 percent) than non-PBS
PBS beneficiaries and 22 percent for non-PBS
female respondents (12 percent).
beneficiaries) or some damage that still
allowed them to live in their homes afterwards.
Only 3 percent of PBS beneficiaries and
2 percent of non-PBS beneficiaries said
their homes were spared from damage during
the storm.
17.
An additional 8 percent of PBS beneficiaries and non-PBS beneficiaries reported having rebuilt their homes with the help of a carpenter.
25
Rebuilding After TC Winston
“More than 70 percent of the homes in Nasaibitu were damaged or destroyed,
this included our village church and village hall.” – Korovou man
“About 90 percent of the houses in Berenitu were partially or fully destroyed.
[The storm] blew off roofs, destroyed our village hall, church, and school.” –
Rakiraki man
“My house was completely destroyed. The wind took the whole roof off, and
my family members were thrown in different directions.” – Rakiraki man
“[TC Winston] ravaged through Tavualevu destroying 70% of our houses,
schools, our village church, and village hall.” – Tavua man
“[The cyclone] left behind homeless people and destroyed most of the Fijian
bures (thatched houses) and a few corrugated iron houses.” – Ba woman
The vast majority of homes of the people interviewed were completely or
partially destroyed by TC Winston. For example, in the Nasaibitu village
(Korovou), 42 out of a total of 66 homes were destroyed – either completely
or partially. Participants attest there are now 80 houses in the village.
Ongoing cash assistance in the form of PBS payments, in addition to Help for
Homes grants and access to free building supplies (provided by government),
enabled most participants to rebuild village homes despite their economic
struggles. Most homes have been repaired or restored to their original condi-
tion, though rebuilding efforts were still underway at the time research was
conducted.
Residents reported having used stronger materials that were more likely to
withstand future storms to rebuild, (such as nuts and bolts instead of roofing
nails) and stronger ropes to secure homes. In Navala Village (Ba), men reported
materials were readily available and free, which helped speed up the rebuilding
process, in addition to cash assistance.
“Just this year we built 12 bures for 12 families...We can now say that at least all
families have a roof over their heads and are comfortable.” – Ba ma
“We have changed the structure of how to build our house so it is suitable
enough if there is a disaster.” – Ba man
Participants in every group reported using cash benefits to buy materials
needed for home repairs. They also admitted that without various forms of as-
sistance (cash benefits and access to building materials) they would have been
unable to upgrade the structural conditions of their homes and instead would
have had to use ‘bush materials’ because they were free.
26
“The benefits I receive every month have helped to buy some small building
materials like nails, paint, and tiles to rebuild my house.” – Korovou woman
“Without these benefits, I do not think I would have been able to rebuild my
house.” – Korovou woman
We are grateful for the assistance the government has given in rebuilding and
repairing our houses. If it was not available then some of us would be living in
makeshift shacks.” – Tavua man.
“Otherwise, we would go back to building thatched houses made of grass or
bush materials.” – Rakiraki man
The Role of Di erent Forms of beneficiaries belonged to households that also
participated in other social protection programs
Social Assistance and Support
(SPS, CPA, DAS), the perception of the
for Immediate and Long-Term importance of social protection increased to
Recovery 69 percent (Figure 9).
The analyses of PBS and non-PBS respondents When asked to rank the most important
suggest that ongoing social assistance is the types of assistance in long-term recovery,
preferred type of assistance for long-term participation in at least one form of social
recovery, for both males and females. The assistance (from SPS, CPA, or DAS) increased
perception of the importance of social the perception of the importance of social
assistance for long-term recovery seems to protection. Of those receiving social assistance,
be influenced by household participation in 54 percent indicated that it was the most
more than one program. Fifty-five percent of important in aiding their long-term recovery
the overall respondents, who only received e orts. On. the. other hand, Help for Homes
social benefits from the PBS, identified social cash grants (62%) were the preferred option of
protection programs as the most important aid the non-PBS beneficiaries who did not receive
in their long-term recovery e orts. When PBS any social protection benefit (Figure 9).
0% 10% 20% 30% 40% 50% 60% 70% 80%
Regular ongoing social assistance
benefits from the Government,
such as the PBS, SPS, CPS, or DIS
Grants from the Help for
Homes program
Other
PBS Only PBS+SPS, CPS, or DIS SPS, CPS, or DIS None
27
Sixty percent of PBS beneficiaries considered non-PBS beneficiaries received F$3,000
ongoing social assistance benefits as the (US$1,434). Thirteen percent of PBS and 14
most important assistance in helping percent of non-PBS beneficiaries reported
restore their general welfare and livelihood figures as high as F$7,000 (US$3,345). Another
in the long term. In kind support, understood 20 percent claimed to have received some
as donations of goods or services, played a other amount through the program.
relatively low contribution to the immediate
Household reconstruction after TC Winston
and long-term recovery but was slightly
was constrained by delays in arrival of
more important for the non-PBS group.
building supplies, sub-optimal materials
It is important to mention that both PBS (not disaster resilient), delays in assistance
beneficiaries and non-PBS beneficiaries and a deficit of skilled human resources,
participated in the Help for Homes program especially in rural areas where there is usu-
in similar numbers and received comparably ally only one builder based in each village.
sized grants. Of the 70 percent who recalled Due to the high post-disaster demand for
receiving grants from the Help for Homes construction labor, many people undertook
program, 48 percent of PBS beneficiaries urgent repairs without seeking assistance from
and 53 percent of non-PBS beneficiaries skilled constructors, and without following
received F$1,500 (US$750). Fourteen percent building codes. It is, therefore, necessary to
of PBS beneficiaries and 12 percent of upskill village-based builders through capaci-
ty-building programs (Aquino et al., 2018).
The Importance of Social Assistance
“These programs are so important to us because they allow our household to
spend the funds as needed to meet our daily needs for our children, family,
village, and church.” – Korovou woman
“These programs are so important because we can buy more food items.” –
Rakiraki woman
“Without these programs, I do not know how we would have survived after
Cyclone Winston.” – Tavua man
All focus group participants were grateful for the ongoing cash they received
from the Government of Fiji. Participants described social assistance payments,
as well as Help for Homes grants, as instrumental in supporting families to
recover their livelihoods. The schemes provided enough for beneficiaries to
survive in the weeks, months, and years since TC Winston hit. Benefits have
been used to pay household bills, for things such as food, childcare, and med-
ical care, to rebuild homes, and to recover losses to businesses and farms.
Multiple participants described benefits as enabling them ‘to put food on the
table’. In addition, some reported having used cash benefits to start small
businesses like canteens, subsistence farms, tailors, and money lending
businesses. Money has also been reinvested in farms that were damaged or
destroyed by TC Winston.
28
“It has helped me to plant my dalo plantation. I must pay three young boys
every month to replant my Dalo plantation that was destroyed by Cyclone
Winston.” – Korovou man
The survey data shows that women were even more reliant on PBS as their
main source of income post-TC Winston. If cash benefits were not available,
participants said they would have no choice but to rely on subsistence farming
to feed their families and to sell whatever little they had leftover.
Women’s Role in Recovery some reporting that they were left out of
recovery e orts. There was also a significant
E orts
decline in paid work opportunities outside of
Focus group discussions provided some the home for women. Social assistance bene-
interesting perspectives on the perceived role fits, therefore, have played a vital role in sup-
of women and men in recovery e orts. Many porting women post-TC Winston.
women felt disempowered, both socially and
economically, following TC Winston, with
Perspectives on the Role of Women
“I had just moved into the village and separated from my husband. I invested
time and money in our taro plantation, and this was all damaged by Winston.
When this happened, I was depressed...I was worried about how we would
survive. It pains me to discuss.” – Korovou woman
“The only obstacles that we women of Nasaibitu are facing is a lack of
education. If there can be short courses provided to us, so we can look for
jobs outside of the village.” – Korovou woman
To some extent, women felt disproportionately a ected by TC Winston and
left out of recovery e orts, with women relegated to more traditional
gender roles within their larger communities and their local economies. The
economic and social disempowerment of women post-TC Winston is evident
in the data.
As mentioned previously, the percentage of respondents working outside the
home declined substantially from pre-TC Winston levels. Just 5 percent of
women were working outside the home at the time of the survey. Focus group
participants described various sources of paid work available to men, like
harvesting sugar cane on neighbouring farms, but reported fewer opportunities
(aside from domestic work) for women post-TC Winston.
29
“There are no kinds of paid work available to us in Korovou or Rakiraki.” –
Korovou woman
One Tavua woman listed sources of paid work available to women as: plucking
out small unwanted weeds from sugarcane fields; peeling cassava; cleaning
mango leaves. For men, she listed: sugar cane harvesting and construction.
Women identified informal sources of income in addition to PBS benefits that
included: sewing; crafting; selling alcohol, suki, loose cigarettes, and frozen
food; and charging other villagers to use their freezer or for electricity to
charge mobile phones. This is in line with the finding that women may be
more reliant on social assistance due to a lack of employment opportunities
available to them. To expand employment opportunities for women, it may be
beneficial to invest in short courses and training focused on providing women
with the necessary skills to find paid work beyond domestic opportunities.
Benefits from social assistance programs, however, have provided a level of
financial independence and stability to women despite disparities in gender
roles. Some women have used PBS payments to invest in their own small busi-
nesses alongside their male counterparts. One woman reported that govern-
ment benefits allowed her to assert a measure of financial independence from
her husband.
“I have started my own small canteen where I sell rice, noodles, kerosene, and
other small stuff.” – Korovou woman
“I have started my own busines, like money lending and tailoring. I have start-
ed selling pillowcases and bedsheets and sell them around the village or in
town to earn some extra cash.” – Tavua woman
“We decide on our own how to use benefits. For example, my husband used
to receive the pension for both of us. If I wanted him to buy something for me,
he wouldn’t give me the money, he would go himself to buy it.
One day I decided to enquire whether I could be part of the program (pension)
as well. Once I was, I did not bother him any more with my needs and wants.
I had my own money to spend and could decide how to spend it.” – Tavua
woman
Many social assistance programs give money directly to women whereas the
PBS is given to the head of the household, which may or may not be a woman.
With this in mind, it may be beneficial to assess the way the PBS is handed out
in order to foster more inclusivity, as long as this can be done in a safe and
culturally sensitive manner.
When asked about village emergency preparedness efforts, villagers attested
that all planning decisions were made by men. Men are in village leadership
30
positions and are responsible for coming up with village emergency plans
when disasters occur.
“Women mainly support any decisions made.” – Ba man
When pressed about the role of men and women in community rebuilding
e orts, focus group participants described the physical role male villagers
played in clearing of debris along the roads, bridges, and rivers and rebuilding
Fijian bures and temporary shelters in the immediate aftermath of TC Winston.
By comparison, male and female villagers characterized women’s roles as
cleaning in and around the house, cooking for men, and caring for children
and the elderly.
Preparedness in the Event of a PBS beneficiaries, there were no di erences
between males and females in their perception
Future Disaster
of being ‘very prepared’ or ‘not prepared at
When the survey went to field in December all’ for a future disaster. Non-PBS beneficiary
2019, a high number (84 percent) of the men (28 percent) were more than twice as likely
respondents believed their communities as non-PBS beneficiary women (12 percent)
would experience another serious climate- to describe themselves as ‘very prepared’.
related natural disaster in the next five Non-PBS beneficiary women were also three
years, including 37 percent who believed it times as likely as their male counterparts to say
was ‘very likely’. Concerns with regards to they were ‘not prepared at all’ (Figure 10).
future disasters were equally high among Additional demographic factors described
both men and women. Results were similar earlier in this report, such as lower rates of
for both male and female respondents of participation in the workforce among female
beneficiary and non-beneficiary groups. respondents, support the finding that women
Seventy-one percent of survey respondents (especially those who do not receive social
tended to believe they were at least ‘somewhat assistance benefits) may be even more
prepared’18 to recover from the e ects of a vulnerable to the negative impacts of a future
future disaster, with 21 percent indicating disaster than their male counterparts.
that they were ‘very prepared’. Among the
Figure 10. Perception of preparedness for a future disaster
70%
60%
Percentages
50%
Very prepared
40%
30% Somewhat prepared
20% Not too prepared
10% Not prepared at all
0% Don't know
PBS female PBS male Non-PBS female Non-PBS male
Female and Male PBS and non-PBS respondents
18.
Qualitative data from FGDs revealed that when participants thought about preparing for a future disaster, they considered making their homes
more secure and resilient, setting aside money for savings, being aware of potential hazards, and stocking up on essential supplies.
31
PBS beneficiaries were more likely to have ties, regardless of gender. Both male and
participated in most of the emergency pre- female respondents in rural communities
paredness activities than non-PBS benefi- reported they would rely on radio communica-
ciaries. Overall, 42 percent of PBS beneficiaries tions in the event of another disaster. Ninety-
and 36 percent of non-PBS beneficiaries one percent said radio was one of their top
received information or training about sources of information. Individuals were much
preparing for another disaster. Furthermore, more likely to seek out information or assis-
39 percent of PBS beneficiary men and 31 tance from family, friends, and neighbours (48
percent of PBS beneficiary women attended percent) or a community leader (16 percent),
Figure 11. Actions taken to prepare for another climate-related disaster
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Received Attended Have Acquired new Bought Put savings aside Purchased
information or community participated in skills to diversify additional to help pay for insurance to help
training about meetings for community income or make emergency unexpected bills protect your
getting prepared disaster activities to build livelihoods less supplies of food, in the event of a family in the
for another response and shelters vulnerable water, first aid or disaster event of a
disaster recovery other tools disaster
PBS Beneficiaries Non-PBS
community meetings for disaster response than they were to turn to police or firefighters
and recovery. These values are 11 and 8 per- (3 percent), or local government (3 percent)
centage points higher than their non-PBS (Figure 12). It is important that governments
counterparts. This indicates the need to use a multitude of channels to reach the
improve access to information and training population in the event of natural disasters,
on disaster preparedness across both groups. even if respondents have not ranked these
channels as top priorities.
Male respondents were more likely than
females to have taken action to prepare for Figure 12. Top sources of information in case
future disasters in the year leading up to of natural disasters
the survey. Within the non-PBS beneficiary
News - Radio 91%
group, 31 percent of male respondents
Family, friends or
reported participating in community activities neighbours 48%
to build shelters, as compared to 15 percent News - Television 45%
of female respondents. Additionally, male Community leader 16%
non-PBS beneficiaries were 10 percentage Newspapers 12%
points more likely to report buying additional News - Internet 6%
emergency supplies than their female Social media 6%
Charitable org.
counterparts. (i.e. Red Cross) 6%
Police or firefighters 3%
The study also found that reliance on radio
Local government 3%
communications in the event of another Religious or cultural
disaster is near universal in rural communi- organization 2%
32
Preparing for Potential Future Disasters
In relation to preparing for potential future disasters, FGD participants explained
that village decision making involved a village headman and clan headsmen.
Most participants – with the exception of villagers from Rakiraki – were unaware
of any communal discussions or the creation of specific village-wide emergency
response plans.
“We usually have a village meeting once a month. Discussions are mainly to do
with the welfare of the village, not so much any risk reduction plans...There
has not been any such committee elected...The key people in the village are
just the Village headsman, the Clan headsman, and the village nurse...They are
all men.” – Ba man
“We have attended community village meetings, which is done every month,
but the discussion of risk reduction development is never part of the agenda.”
– Tavua woman
The village of Rakiraki seems further along in disaster planning than other
villages; villagers who participated in focus groups were actively planning for
future disasters. Rakiraki participants reported having appointed a disaster
response committee responsible for village planning, though the committee
had not yet convened at the time focus group discussions were held.
“In my village clan, we have only begun to organize such meetings and that is
to happen next week. We have invited someone to provide advice on what
measures to take in the event of a natural disaster.” – Rakiraki woman
“Winston really taught us a huge lesson in the need to become more prepared
and aware.” – Rakiraki man
As the data indicates, levels of personal preparedness varied across households.
Many villagers rebuilt their homes using materials better able to withstand
future storms. Some villagers (Rakiraki) attested to stocking up on supplies like
food, water, and clothing supplies in the event another disaster occurs or
putting aside money in savings.
“There is no plan since there is no committee for a natural disaster, but as
individuals we know what to do.” – Tavua man
“Preparation and plans are done by individual households and that is to invest
more and keep more money aside if there is another disaster to help with a
speedy recovery.” – Ba woman
Participants uniformly requested more specific and clearer information about
emergency preparedness and what to do in the event of another natural disaster,
such as greater advance warning about approaching storms with advice about
where to go and what to do to ride out the storm safely.
33
“We were not prepared. We received the information late. The cyclone reached
us before the warning came.” – Rakiraki man
“Most of us are uneducated, and we do not know the meaning of ‘Category 5’.
We are only used to the knots of the wind.” – Korovou man
34
3. KEY FINDINGS
This study identified that Fiji’s social protection • A significant number of low-income Fijians
system has provided important assistance to described themselves as worse-o finan-
recipient families in managing the long road cially than they were before TC Winston
to economic recovery following TC Winston, occurred. Roughly half (47 percent) of
which wiped out livelihoods and infrastruc- those surveyed in a ected communities
ture and heavily damaged or destroyed the vast reported that they had less money (29 per-
majority of homes in a ected communities. cent ‘somewhat less’ and 18 percent ‘much
less’) to pay for things like food, housing,
The findings of the study, conducted in 2019,
and electricity than they did before TC
are summarised below:
Winston. Employment opportunities declined
• Ongoing social assistance benefits have following TC Winston, particularly for
been important for household resilience vulnerable groups, and the role of social
and recovery from large shocks such as assistance as the main source of income,
TC Winston and other household shocks. supporting income shocks, has increased.
The report found that perception of the im- The importance of the PBS as households’
portance of social assistance on long-term main source of income quadrupled from
recovery seemed to be influenced by 2016 to 2019.
household participation in more than one
• Women recipients (26 percent) were twice
program. In addition, Help for Homes cash
as likely as men (13 percent) to have re-
grants (62 percent) were the preferred
ported the PBS as their main source of
option of the non-PBS beneficiaries
household income. Only 5 percent of
interviewed who did not receive any social
female respondents reported having em-
protection benefits.
ployment outside of the home post-TC
• Despite having made significant progress Winston. This may be due to the fact that
in rebuilding their personal economic rates of employment, which were already
security, local infrastructure, and homes, lower for female respondents as compared
those surveyed reported continued eco- to their male counterparts, fell by 66 per-
nomic hardship. Both PBS beneficiaries cent for female PBS beneficiaries. Lack of
and non-PBS beneficiaries (the near poor) access to childcare poses a constraint to
described themselves as experiencing women’s participation and re-entry into
economic hardship (that is, unable to meet the workforce. Some women, nonetheless,
their basic food and non-food needs), with have reported using social assistance
86 percent of PBS beneficiaries and 83 percent benefits to invest in business ventures,
of non-PBS beneficiaries saying they were providing a level of financial independence
among those experiencing the highest levels and income despite disparities in gender
of hardship compared to the rest of their roles and employment opportunities.
village.
35
• Many poor and near poor Fijians interviewed • Ongoing cash assistance in the form of
remain disproportionately vulnerable to PBS payments, in addition to Help for
the e ect of another serious disaster, Homes grants and access to free building
with 84 percent of respondents believing supplies (provided by government),
they personally will be seriously a ected by enabled most participants to rebuild village
another natural disaster or storm in the next homes despite their economic struggles.
five years, including 37 percent who believe Furthermore, FGD participants reported
it is ‘very likely’. Neither PBS nor non-PBS using social assistance benefits to build
beneficiaries indicated feeling prepared for back better, with stronger materials that are
future disasters or shocks. Based on their more likely to withstand future storms,
past experiences, only 21 percent of PBS although there seems to have been issues
beneficiaries perceived their household with sub-optimal materials and deficit
to be ‘very prepared’ to recover within a of skilled human resources. One-third of
reasonable time from a future shock or respondents whose homes were damaged,
disaster. Non-PBS male beneficiaries however, have been unable to finish repai-
(28 percent) were more than twice as likely ring all of the damages incurred by TC
as non-PBS female beneficiaries (12 percent) Winston (31 percent PBS beneficiaries,
to describe themselves as ‘very prepared’. 30 percent nonPBS).
Non-PBS female beneficiaries were also
• PBS benefits were used to pay for essential
three times as likely as their male counter-
households needs, with food expenses
parts to say they were ‘not prepared at all’.
comprising the bulk of benefit expendi-
This highlights the importance of providing
tures. More than 90 percent of PBS
training and support to prepare and respond
beneficiaries reported to primarily using
to future shocks.
monthly benefits to pay for food. Benefits
remaining after paying for food were
largely used to pay for bills like phone, gas,
electricity, and health-related expenses.
Only 5 percent of respondents spent a
portion of benefits on social obligations.
There was no di erence in reported
expenditure by male and female respon-
dents. Non-PBS beneficiaries had similar
expenditure patterns, with most of their
expenditure spent in food and paying bills.
36
4. RECOMMENDATIONS
The following is a set of recommendations the Recommendation 3: Assisting vulnerable
Government could consider to improve social households to diversify their income through
assistance programs in terms of targeting/- access to training and finance could be
coverage of beneficiaries and vulnerable achieved by establishing linkages and referral
households; delivery system strengthening; systems to connect beneficiaries of Fijian
referral and linkages for income diversification; social protection programs with other service
and disaster preparedness and training - all providers in their regions such as employment
aligning with social protection policy reforms programs and financial inclusion services.
and with disaster response coordination e orts. Linkages and referral mechanisms maximize
the impact of investments in di erent sectors,
Recommendation 1: The government could
consider building and improving existing putting together service providers and benefi-
systems to facilitate easy expansion of ciaries, broadening social protection programs
social assistance programs (both vertically to specifically address more vulnerabilities. This
and horizontally) to provide predicted is particularly important for female beneficia-
support to vulnerable non-PBS beneficiaries ries who were found to have less access to paid
and top-ups to PBS beneficiaries. For work outside the home. The report supports
example, the government could consider the IFC recommendation of looking into the
establishing a Social Registry to better help issue of childcare in Fiji, considering access,
identify the poor and vulnerable in times of a ordability, regulatory framework, safety,
disaster. Having systems that are interoperable and quality.
and able to quickly identify vulnerable Recommendation 4: Ensure groups that are
populations through overlay with disaster particularly vulnerable to disasters (such as
information provide timely support after a the PBS and near poor households, people
disaster.20 living with disabilities, single mothers, older
Recommendation 2: Further analysis is required persons, and others) have timely access to
to consider how social protection can be information, training and support to prepare
tailored (in normal times and in response to and respond to future shocks, and to develop
covariate shocks) to address specific risks, and implement disaster risk reduction plans in
vulnerabilities, and inequalities for women and collaboration with other community members
marginalised groups, in order to improve their and government agencies to address specific
resilience to disasters.21 hazards that the community will face.
19.
The World Bank supported a business process review of the Department of Social Welfare social assistance programs identifying areas for system
improvement. Please refer to Annex 2 for more information.
20.
In 2021, the Social Assistance Policy has included the development of a social registry, the update of the MIS and the updated SOPs as part of its
priority actions.
21 .
As part of the Fiji Social Protection COVID-19 Response and Systems Development Project, the Government of Fiji will be developing an Adaptive
Social Protection Strategy, as a whole-of-government e ort to quickly respond to natural disasters and other covariate shocks. The World Bank will
provide technical assistance to the Government including gender-smart and adaptable social protection programs.
37
Given that participants across all focus groups Recommendations 5: Identify opportunities
highlighted the need for more specific and to support women and improve the inclusivity
timely information regarding future disasters of disaster risk reduction programming
(for example, clarifying the meaning of storm to involve women as agents of change.
categories, describing storms in knots, and Given that women often plan an important
providing greater advance warning about role in preparing for risk and acting as first
approaching storms), it may be beneficial to responders but have unequal participation in
involve community members in program disaster response planning, it is important that
design to ensure training terminology is acces- support be specifically targeted at encouraging
sible and relevant. Additionally, supplementary women to participate in planning meetings
disaster risk reduction programming and and helping them stay engaged throughout
resources could be provided during existing the recovery phase, particularly in the deci-
community meetings to boost village-wide sion-making processes. Gender responsive
participation. disaster risk reduction frameworks22 should
be used as part of capa city building to
increase advocacy within the community
for establishing gender responsive actions
(including through advice on climate resilient
livelihoods, early warning systems, disaster
insurance, women’s disaster preparedness
networks, and women- led community
knowledge sharing).
22
Disasters a ect women, boys, girls, and men in di erent ways. By taking into account needs, opportunities, roles, relationships, and gender
norms – the gender-responsive disaster risk reduction framework is a pro-active and people-centered approach to reducing vulnerabilities and
risks in the event of a disaster.
38
ANNEXES
Annex 1.
Social Protection Scheme
Scheme Target group Benefits (monthly) Possible overlap
Poverty Benefit Households that fall below F$35 per adult and F$17 per
Scheme (PBS) the defined poverty threshold child (or else any variation that
(targeted bottom 10%) at the will give a max. of F$127).
bottom 10%)
Apart from monthly allowance,
• Will assist up to 4 members an additional F$50 food
in each household voucher is given to each
recipient household.
Social Pension Elderly persons who are at F$100 per recipient Can be accumulated with
Scheme least 65 years of age with no the CPA in the case the
source of income and not elderly person is the
receiving government primary care giver for a
pension, ex- serviceman grandchild/child until
funds or FNPF pension they reach the age of 18.
Care and Protection Those living with Min. amount: F$29
Allowance (CPA) permanent disabilities. Max. amount (regardless of the
number of children): F$127
• Pre-school: F$29
• Primary school: F$35
• Secondary school: F$46
• Child with disability: F$69
An additional F$50 food
voucher is given to each
recipient household.
Disability Allowance F$90 per person for a max.
Scheme (DAS) of 2 household members (max.
allowance per household:
F$180)
39
Annex 2. households and individuals, contributing to
poverty alleviation as well as the empowerment
Business Process Review (BPR)
of vulnerable and disabled individuals. DSW
programs fall under the Social Policy Act which
The World Bank supported a business process
was approved in 2001 and expired in 2011.23
review (BPR) of Department of Social Welfare
The DSW uses a proxy means test (PMT)24 for
(DSW) social assistance programs including
targeting the poorest population, while the
the PBS, CPA SPS and DAS, as requested by
elderly and disability programs are categorically
DSW. The process involved a two-day workshop
targeted. The PMT is embedded in the applica-
with 20 attendees including management and
tion form to determine poverty eligibility. In
field o cers of DSW. The workshop reviewed
addition, the MIS is decentralized to social
the processes of the five programs, where the
welfare o ces (around 17), but the data cannot
DSW operational sta from the central and dis-
be consolidated at the central level, thus the
tricts engaged actively and provided detailed
DSW does not have a centralized MIS.
inputs on the operational challenges of the
programs. The BPR consultation document DSW business processes require attention
outlines the challenges of current operations and there is an opportunity to strengthen the
and processes, and which improvements can system resilience to adapt to crises so that
be made with new integrated MIS. These im- systems and processes are prepared to adapt
provements aim to increase resilience by deve- and deliver services in a timely manner in the
loping more adaptive and shock-responsive event of a future disaster. Fiji DSW programs
social protection systems that can respond in a were revamped with the aim of improving their
timely manner. The BPR examined the existing poverty-targeting and e ectiveness in 2012,
business processes of DSW programs captu- with assistance from the World Bank. DSW has
ring every step performed by the DSW o cers made constant e orts to improve their pro-
for the programs. The final BPR consultation cesses. However, the MIS remains an adminis-
document also proposed processes to be trative and strategic management challenge
implemented to streamline and standardize for DSW. The process review found that the
DSW programs when new MIS are introduced. existing MIS is fragmented and lacks automa-
Thus, the BPR could provide the functional tion, several databases are managed at the
requirements for developing a new integrated district o ce level and therefore lacking an
MIS for DSW. The final document and MIS integrated (national) MIS, and the existing SOPs
proposal were completed in January 2020 and are outdated.
presented to the Ministry of Economy in Fiji. A key finding of the BPR was that policies for
The DSW maintains MIS for several social social assistance programs and the SOPs are
protection programs: the PBS; the SPS; the outdated and not widely utilized by sta and
CPA; the DAS – along with two additional therefore do not seem to be providing value
programs, the Food Voucher and the Bus to their day-to-day operations. This is hinde-
Voucher. Through these programs the DSW ring management of the schemes. It also found
provides supports to approximately 87,000 that there are misunderstandings about DSW
23.
A new Social Assistance Policy was approved by the Government of Fiji in 2021.
24.
The proxy means test (PMT) is a welfare-based methodology used to estimate income based on household (HH) consumption and assets.
Coe cients are calculated based on the HIES data using a statistical model which provides a score to select poor beneficiaries for the PBS.
40
programs within the DSW. These processes for participation in the survey or focus groups.
should be clearly reflected in updated SOPs.
Development partners continue to provide The Survey
support to DSW to address some of the outlined Tebbutt Research’s Fiji-based research team
issues.
conducted 446 interviews with PBS and
Annex 3. non-PBS beneficiaries living in Ba, Rakiraki,
Tavua, and Korovou in December 2019.
Methodology of the Survey The sample was stratified by province to
and Focus Group Discussions ensure the sample’s regional distribution was
To achieve the desired outcomes, the World proportional to population statistics for the
Bank, with guidance from third party consul- target groups. The sample was further stratified
tants, designed a mix method study consisting to ensure a roughly equal proportion of male
of a household survey, followed by qualitative and female respondents, PBS beneficiaries
focus group discussions. and non-PBS beneficiaries within each of
the four regions sampled. Tebbutt Research
The household survey and qualitative focus conducted an additional 46 interviews to the
groups were conducted in the regions of Ba originally proposed sample of 400 completed
(Western Division), Rakiraki (Western Division), interviews resulting in a slightly higher number
Tavua (Western Division), and Korovou of PBS beneficiaries than non-PBS beneficiaries
(Central Division), which encompassed the in the final sample. The margin of error is
majority of TC Winston-a ected households. +/-6% at the 95 percent confidence level
The study sample was divided into two groups, for both the PBS beneficiary and non-PBS
consisting of: (i) individuals who received beneficiary samples.
the Poverty Benefit Scheme (PBS) from the A random sample of potential survey respon-
Government of Fiji before TC Winston dents was generated from a master list
occurred (through to 2019) and (ii) those provided by the Fiji Department of Social
members of the near poor, or individuals from Welfare (DSW). The list provided by DSW was
low-income households who applied for the comprised of beneficiaries of the Poverty
PBS but whose applications were declined Benefit Scheme (PBS beneficiaries) as well as
by DSW because they slightly exceeded the individuals who applied for but whose appli-
cut-o point of the PMT. cations for the PBS were declined (referred to
The survey sample was delineated to determine throughout this report as non-PBS beneficia-
whether there were di erences in recovery ries). DSW only provided contact information
and resilience over time between households for PBS beneficiaries who were enrolled in the
that received the PBS and those who did not, PBS before TC Winston occurred (as far back
taking a long-term look at the role of cash as 2013) and who were still enrolled in the PBS
transfers beyond the immediate aftermath of at the time the survey was conducted. The list
TC Winston. Consideration was also made as of non-PBS beneficiaries provided by DSW
to whether PBS and non-PBS beneficiary was comprised of individuals who have never
households might have members enrolled in qualified for nor received PBS payments.
other social protection programs. Non-PBS beneficiaries were defined by DSW
as individuals living near or just above the
All respondents must have been living in the
poverty level as per the PMT (the near poor)
a ected areas when TC Winston hit to qualify
41
but whose requests for benefits were denied ries; with one group conducted with male and
for failure to meet the full set of criteria used one group conducted with female beneficia-
to determine eligibility. The concept of proxy
means testing is usually understood as using focus group discussion accommodated be-
observable characteristics of the household tween 8 and 10 participants. Focus group dis-
or its members to estimate their incomes, cussions were hosted in the following villages:
based on consumption, when other income
• Navala Village, Ba
data (salary slips, tax returns) are unavailable
or unreliable. As the name suggests, the • Nasaibitu Village, Korovou
obtained income estimates are in turn used • Burenitu Village, Rakiraki
for means-testing, that is to determine whether
• Tavua Levu Village, Tavua
a household, or an individual, is eligible for
support from a social protection program. If Whereas research audiences were chosen in
their estimated income falls below a certain relation to their participation with the PBS
threshold, they are deemed eligible for support; specifically, focus group participants tended
if the income is higher than the threshold, to discuss their experiences with ongoing
they are ineligible. Eligibility is reviewed every social assistance programs more generally,
three years. Non-PBS beneficiaries were including the PBS, SPS, CPA, DAS, or food
defined by DSW as individuals living near or voucher programs. They did not distinguish
just above this poverty threshold, having
therefore the comparable characteristics with responses.
the PBS beneficiary group. All data was
de-identified to ensure the highest levels of
data privacy and respondent confidentiality.25
Focus Group Discussions
Follow-up focus group discussions were
conducted exclusively among PBS beneficiaries
in the regions surveyed in order to explore
experiences with social assistance benefits
and the role of ongoing cash transfers in the
recovery process at a more nuanced level
than close-ended survey questions allow.
Tebbutt Research conducted a total of eight
focus group discussions with PBS beneficia-
25.
The report followed a systematic approach in the sampling selection to demonstrate a higher likelihood of comparison between groups that share
similar characteristics based on the selection of survey respondents around the PMT cut-o point. However, it is important to note that this is a
qualitative report and the groups were not identified following an experimental design of treatment and control groups, required to measure impact
of the intervention.
42
Survey Sample Frame
DSW LIST TARGET SAMPLE
Province DSW PBS Non-PBS Percentage Target PBS Non-PBS
DSW LIST
list beneficiaries beneficiaries by district TARGET SAMPLE
Sample size beneficiaries beneficiaries
(n=400) (50% male, (50% male,
Province DSW PBS Non-PBS Percentage Target PBS female)
50% Non-PBS
50% female)
list beneficiaries beneficiaries by district Sample size beneficiaries beneficiaries
(n=400) (50% male, (50% male,
Ba 2288 1390 898 46% 184 92 92
50% female) 50% female)
Korovou 529 446 84 11% 44 22 22
Ba
Tavua 2288
1185 1390
574 898
611 46%
24% 184
96 92
48 92
48
Korovou
Rakiraki 529
948 446
791 84
157 11%
19% 44
76 22
38 22
38
Tavua
TOTAL 1185
4950 574
3201 611
1750 24%
100% 96
400 48
200 48
200
Rakiraki 948 791 157 19% 76 38 38
TOTAL 4950 3201 1750 100% 400 200 200
FINAL SURVEY SAMPLE
Province Total N PBS beneficiaries Non-PBS beneficiaries
FINAL SURVEY SAMPLE
Ba 197 98 99
Rakiraki
Province 78
Total N 39
PBS beneficiaries 39
Non-PBS beneficiaries
Tavua
Ba 118
197 70
98 48
99
Korovou
Rakiraki 53
78 30
39 23
39
Tavua 118 70 48
Male respondents
Korovou 226
53 119
30 107
23
Female respondents 220 118 102
TOTAL
Male respondents 446
226 237
119 209
107
Female respondents 220 118 102
TOTAL 446 237 209
Summary Statistics PBS Non-PBS
Characteristics Option Beneficiaries beneficiaries Di erence p-value
Age Number PBS
54.9 Non-PBS
53.8 1.1 0.27
Characteristics Option Beneficiaries beneficiaries Di erence p-value
Primary 51.9% 49.3% 2.6% 0.58
Education level
Age Secondary
Number 40.5%
54.9 46.4%
53.8 -5.9%
1.1 0.21
0.27
Married
Primary 60.8%
51.9% 63.2%
49.3% -2.4%
2.6% 0.6
0.58
Marital status
Education level Widow/er
Secondary 29.5%
40.5% 26.8%
46.4% 2.7%
-5.9% 0.52
0.21
Household size Number
Married 3.9
60.8% 4
63.2% -0.1
-2.4% 0.62
0.6
Marital
No. status in HH
of females Widow/er
Number 29.5%
2.00 26.8%
2.11 2.7%
-0.11 0.52
0.39
Household
No. of malessize
in HH Number 3.9
1.98 4
1.97 -0.1
0.01 0.62
0.91
No. of
No. children in
of females in HH
HH Number 1.14
2.00 1.38
2.11 -0.245
-0.11 0.98
0.39
Location
No. of males in HH Urban
Number 11.8%
1.98 10.1%
1.97 1.7%
0.01 0.55
0.91
No. of children in HH Rural
Number 88.2%
1.14 89.9%
1.38 -1.7%
-0.245 0.55
0.98
Location Urban 11.8% 10.1% 1.7% 0.55
Rural 88.2% 89.9% -1.7% 0.55
43
Note on the Analysis • In FGDs, some PBS beneficiaries were
unable to clearly distinguish between
Significance refers to 90 percent and 95 percent
benefits their household received from
confidence levels.
other schemes and PBS benefits, as cash is
• Significant Di erences at the 95% CI (critical fungible. This may confound some findings
score Z score <-1.96/ >+1.96, p<0.05) related to the use of PBS benefits as opposed
• 90% CI (critical Z score <-1.65/>+1.65), to other social assistance benefits the
p<0.1) household may have been receiving.
• Given the time that has lapsed since TC
Study Limitations Winston, many respondents had issues with
• Survey captures PBS and non-PBS benefi- recall at the time of the survey and FGDs.
ciaries also involved in other programs; this
represents 35 percent of the sample size Annex 4. Areas for Further
for the PBS group and 41 percent for the Research
non-PBS group and they are comparable in
• Research on the impacts of other programs
characteristics. It is unknown how partici-
(DAS, SPS, CPA) on disaster recovery.
pation in other schemes could influence
responses. • Review of the social assistance benefit levels.
• Given that DSW only provided contact • Additional opportunities to undertake impact
information for PBS beneficiaries who were evaluation on programs, particularly with
enrolled prior to TC Winston and remained renewed changes in policy and programs
enrolled at the time of the survey, those within DSW. This could include evaluating
who may have graduated out of the program access to health, education, and social
or those who qualified for the benefit services, including rates of school attendance
post-TC Winston were not included in the for households receiving benefits and how
sample. This has potential implications for the benefits a ect work incentives.
the comparability of the PBS and non-PBS • Given qualitative data from FGDs in which a
beneficiary groups in terms of the impact very small number of women have expressed
TC Winston may have had on a household’s the need to request their own access to
economic situation. Furthermore, this has PBS – a further analysis to understand how
the potential to skew the results downwards female targeted programs could be imple-
as those who were unable to graduate from mented as part of a gender-sensitive social
the program may have been those who assistance program design.
would have had less opportunities to find
alternative sources of household income.
44
REFERENCES
Asante, A. D., Irava, W., Limwattananon, S., Hayen, Mansur, A., Doyle, J., and Ivaschenko, O. (2018).
A., Martins, J., Guinness, L., Ataguba, J. E., Price, Cash Transfers for Disaster Response: Lessons
J., Jan, S., Mills, A., and Wiseman, V. (2017). from Tropical Cyclone Winston. Development
Financing for universal health coverage in small Policy Centre Discussion Paper No. 67, Available
island states: evidence from the Fiji Islands. at http://dx.doi.org/10.2139/ssrn.3143459
BMJ global health, 2(2), e000200. https://doi.
National Disaster Management O ce. (2016).
org/10.1136/bmjgh-2016-000200
Tropical Cyclone Winston Situation Report 95.
Aquino, H.D., Wilkinson, S.J. Raftery, G.,
World Bank Group. (2020). Country Partnership
Potangaroa, R., and Chang-Richards, A. (2018).
Framework for the Republic of Fiji FY2021-
Challenges to building housing resilience: the
FY2024.
case of Fiji post-cyclone Winston. Procedia
Engineering 212. 475–480. World Bank Group. (2021). Fiji - Social
Protection COVID-19 Response and System
Burton, C. (2015). Fiji: Making Social Protection
Development Project - Project Appraisal
More Responsive To Natural Disasters and
Document. Washington, D.C.: World Bank
Climate Change.
Group.
Dornan, M.D, Silva Parsons, K.H., Doan, D.T., Yi,
World Bank Group. (2022). East Asia and Pacific
S., Arahan, R.J., Munoz, A.G., and Vergara Hegi,
Economic Update April 2022: Braving the
D.E. (2020). Pacific Island Countries in the Era
Storms. Washington, D.C.: World Bank Group.
of COVID-19: Macroeconomic Impacts and
Job Prospects (English). Washington, D.C.:
World Bank Group.
IFC. (2019). Tacking Childcare: The Business
case for Employer-Supported Childcare in Fiji.
Mansur, A., Doyle, J., and Ivashenko, O. (2017).
Social Protection and Humanitarian Assistance
Nexus for Disaster Response: Lessons Learnt
from Fiji’s Tropical Cyclone Winston. World
Bank, Social Protection and Labor, Discussion
Paper Number 1701.
45